This article is so true. "Collaboration" is how nothing ever gets done; we have this expression: "designed by committee"; we should also have "made by collaboration".

What's depressing is that it's like Fred Books' book never happened: most managers think the way to solve IT problems is just to trow more people / more money at it until it gets solved; and they're all surprised when it doesn't work, but try again the next time anyhow.

I think "design by committee" is a better target for criticism than collaboration in general.

If you get a bunch of people in a room and ask them for a design, one person is going to write the design while everyone else gets in the way. That's simply the nature of groups. The one person who writes it isn't even necessarily the best designer—they're just the one most willing to grab the whiteboard marker.

Conversely, if you ask one person to produce a preliminary design, they can leave, gather requirements, do research, produce a plan, and then convene everyone in a room to review it. Now all the abstract hypotheticals have been put to bed, the nebulous directionlessness has been replaced with a proposal, and the group can actually provide useful feedback and have a discussion that will inform the next draft of the design. And once the design is finished, everyone can easily work together to implement it as written. Collaboration is great, after someone has proposed a design.

That's part of what I like about the idea of Amazon's "culture of writing," though I've never worked in an environment like that in practice. Every idea needs to be preprocessed into an actionable memo before anyone tries to have a meeting about it.

Lets compare two projects that are collaborations:

Linux and Wayland.

Both are collaborative efforts, one has fairly effective and tyrannical leadership with the best interests of the community in mind. The other is lead by committee with competing interests and goals where they all have veto power.

Those same collaborators are reflected in the distro situation... Here is a group that also has some rather tyrannical leadership but they have dependencies (see the software they run) and some of those folks are sick of the distro's maintainers nonsense, and went to things like flat packs (see Bottles for an example).

> most managers think the way to solve IT problems is just to trow more people / more money

Leadership vs Management, a tale as old as time.

Right. But more often than not, the problem that's being solved is "we have gotten money to throw at things", so the answer of throwing in many more people to busywork kind of makes sense.

That's before we even think about all the consultants and similar roles where busywork really is work. Then all the organizational or agile roles.

The fact that some product gets shipped and we still have customers is good, because that's what pays for it all, but that is just the foundation we all rest on. Almost like background noise.